Theme: Environment, Health and Economics EERC Working Paper Series: EHE-2

Economic Valuation of Health Damage in North Chennai Using a Comparative Risk Assessment Framework

Kalpana Balakrishnan

Ramachandra Medical College and Research Centre, Chennai

MOEF IGIDR WORLD BANK

EERC

1

ECONOMIC VALUATION OF HEALTH DAMAGE IN NORTH CHENNAI

USING A COMPARATIVE RISK ASSESSMENT

FRAMEWORK

Project Report

Submitted by

SRI RAMACHANDRA MEDICAL COLLEGE AND RESEARCH INSTITUTE (Deemed University)

To

THE INDIRA GANDHI INSTITUTE FOR DEVELOPMENT RESEARCH

Under

The World Bank Technical Assistance Program On

Capacity Building in Environmental Economics

March 2001

2

CONTENTS

Acknowledgements -------------------------------------------------------------------------------------- 3 Executive Summary -------------------------------------------------------------------------------------- 8 Background ------------------------------------------------------------------------------------------------- 16 Objectives --------------------------------------------------------------------------------------------------- 22 Methodology ------------------------------------------------------------------------------------------------ 23 Collection of secondary environmental data Collection of primary environmental data Collection of secondary health data Administration of health assessment questionnaire Collection of secondary socioeconomic information Collection of primary information on health expenditure Health risk assessment using established dose –response information

Health impact assessment using cross sectional epidemiological information Economic valuation of health impacts

Results ------------------------------------------------------------------------------------------------------- 35

Description of study area Socioeconomic profile

Environmental profile Environmental quality Health Impacts

Economic valuation Ranking of environmental concerns--------------------------------------------------------------- 49 Conclusions ------------------------------------------------------------------------------------------------ 50 Recommendations --------------------------------------------------------------------------------------- 51 References ------------------------------------------------------------------------------------------------- 54 Annexures Field Questionnaires

3

ACKNOWLEDGEMENTS

The investigators would like to thank and express their gratitude to a many individuals and

organisations without whose co-operation this study would not have been possible.

We are grateful to our Vice- Chancellor, Dr. T.K. Partha Sarathy , whose support throughout

the course of the study was invaluable. We also sincerely thank our Director, Academic

Administration, Dr. D. Gnanaprakasam and Dean of Faculties, Dr. S. Thanikachalam, who

ensured the mobilisation of staff members from various departments for the execution of the

project.

Several staff members of SRMC&RI lent their support. In particular, we wish to thank Dr. K.V.

Somasundaram, Head of the Department of Medicine, Dr. Vijayalakshmi Thanasekaran,

Head of the Department of Chest Medicine, Dr. Brig.. V. Satyamurthy, Head of the

Department of Community Medicine and Dr. B. W. C. Satyasekharan, Professor of

Community Medicine, for their valuable guidance.

The Tamil Nadu Pollution Control Board has been actively involved during the entire course of

the study. We thank, Mrs. Siela Rani Chunkath, I.A.S., Chairperson, TNPCB, Shri. M.

Devaraj, I. A. S., Former Chairman, Tamil Nadu Pollution Control Board (TNPCB), Mr.

Rangasamy, Member secretary, TNPCB, Dr. K. Mani, Deputy Director, AEL, TNPCB and Dr.

V.N. Rayudu, Director, ETI, TNPCB for their valuable support and technical assistance. The

sampling equipment loaned for the project by TNPCB is gratefully acknowledged.

The study would have been impossible without the co-operation of the officials of the

Thiruvootiyur Municpality. In particular we wish to thank Mr. T. V. Vishwanathan, Chairman ,

Mr. T. Gopinath, Commissioner and Mr. V. Vasavakumar, Medical Health Officer of

Thiruvottiyur Municipality for their support.

Our special thanks to Dr. Paul Appasamy, Director, Madras School of Economics for his

continued guidance during the project. Several members form the Indira Gandhi Institute for

Development Research, Mumbai(IGIDR), helped us through numerous informal discussions

that were crucula in steering the project. We thank Dr. Kirit Parikh, Former Director of IGIDR,

Dr. (Mrs.) Jyothi Parikh and Dr. (Mrs.)Vijaylakshmi Pandey for their technical involvement.

4

We also thank Dr. P. Venkatesan, ICMR, Dr. Sharadha Suresh, Madras Medical College and

Mr. D. Dharmarajan, Madras Institute of Development Studies for their assistance in statistical

analysis.

Finally, we thank our Chancellor, Mr. V. R. Venkatachalam and our management who have

always stood behind each one of us in all our research endeavours.

Staff of Environmental Health Engineering Cell

Sri Ramachandra Medical College & Research Institute (DU)

5

LIST OF INVESTIGATORS FROM SRMC & RI

Principal Investigator

Dr. Kalpana Balakrishnan, Environmental Health Engineering Cell

Environmental Quality Team Members

Mr. S. Shankar Environmental Health Engineering Cell Dr. K. Srividya Environmental Health Engineering Cell Dr. Vidhya Venugopal Environmental Health Engineering Cell Ms. Swarna Prasad Environmental Health Engineering Cell Mr. D. Venkatesh Environmental Health Engineering Cell

Health assessment Team Members

Dr. V. Ramasubramanian, Department of Medicine Ms. S. Ramani, Staff Nurse Ms. V. Shyamala, Staff Nurse Ms. T. Velvizhi, Staff Nurse Mr. P. Venkatesan, Staff Nurse Mr. K. Venkatesan, Ward Technician

Economic Assessment Team Members

Mr. Narahari Reddy Environmental Health Engineering Cell Mr. S. Devanthan Environmental Health Engineering Cell

Advisory Team

Dr. K. V. Somasundaram, Head, Department of Medicine Dr. Vijayalakshmi Thanasekharan, Head, Department of Chest Medicine Dr. Brig. V. Satyamurthy, Head, Department of Community Medicine Dr. B.W.C. Satyasekaran, Professor, Department of Community Medicine

MEMBERS FROM OTHER INSTITUTIONS INVOLVED IN THE STUDY

Advisory Committee

Chairperson Mrs. Shiela Rani Chunkath, I.A.S.,

Chairperson TNPCB Chennai 600025 Project Advisor Dr. Paul Appasamy Director Madras School of Econmics Chennai 600020 Members Dr. T. Ramasami Dr. Jai Prakash Narayan Director Director, CES CLRI Anna University Chennai 600025 Chennai 600025 Dr. Y.V. Subramaniam Mr. A. Doss Scientist-in-charge Chief Urban Planner NEERI Unit, ChZL CMDA

Taramani Egmore Chennai 600113 Chennai 600008

Mr. R. Ramanathan Dr. V. N. Rayudu S. E. (SWM) Director Corporation of Chennai Environmental Training Institute Chennai 600002 TNPCB Chennai 600025

Working Groups Statistics 1. Dr. Sharada Suresh Chief of Epidemiology Madras Medical College Chennai 2. Dr. Nagaraj

Professor of Statistics MIDS, Chennai

3. Dr. P. Venkatesan Senior Research Officer Tuberculosis Research Center ICMR, Chennai.

7

Environmental Data Collection 1. Dr. Jayabalou 4. Dr. P. Chandrasekar Scientist, NEERI R & D in-charge Chennai Zonal Laboratory TNPCB Chennai Chennai 2. Mr. Charles Rodriguez 5. Mrs Indira Gandhi TNPCB Ambattur office Chennai TNPCB, Chennai 3. Mr. Nagaraj Geochemist Ground Water Cell PWD Chennai Economic Valuation 1. Dr. G. Mythili Madras School of Economics Chennai 2. Dr. Paul Appasamy Director, Madras School of Economics Chennai Field/ Health Data Collection

1. Mr. T. V. Viswanathan

Chairman Thiruvottiyur Municipality

2. Mr. Gopinath

Commissioner Thiruvottiyur Municipality

3. Dr. Rasakumar

Medical Health Officer Thiruvottiyur Municipality

4. Dr. M. S. Vijayakumari Medical Health Officer Kuppam Health Post Ramanathapuram

5. Dr. S. Andal Medical Health Officer Thangal Health Post

6. Dr. Mani Carmel Officer Sathangadu Heath Post 7. Dr. M. Devika Medical Health Officer Thiruvottiyur Health Post 8. Dr. Saraswathy Medical Health Officer Eranavoor Health Post

Executive Summary

Introduction

The general objective of the project was to rank a set of environmental problems faced by a

municipality within an industrial zone of Chennai City using a comparative risk assessment

framework (CRA). The CRA approach originally developed by the United States

Environmental Protection Agency (USEPA), has been applied for aiding resource allocation

for environmental management in several states of USA as well as in many cities of

developing countries including Ahmedabad in India and Asansol in Bangladesh. One of the

biggest impediments to successful application of the CRA framework in developing countries

is the lack of both quantitative exposure and health effects information. The project

demonstrates the usefulness of the CRA approach by addressing many of these deficiencies

through rigorous primary and secondary data collection on environmental, health and

economic parameters. The study represents one of the first studies in India to use a

quantitative health risk assessment framework to address environmental health concerns at a

municipal level with direct stakeholder involvement.

Objectives

The focus of the project was to assess human health risks associated with particular

environmental exposures using both previously established dose – response information and

cross-sectional epidemiological information collected during the course of the study. The

economic costs associated with these health risks were then evaluated using local economic

information to enable ranking of the particular environmental concerns on the basis of both

the health and economic risks. The geographic focus of the project was the Thiruvottiyur

Municipality in North Chennai. Since even within a single zone, there are a multitude of

environmental concerns, the project sought to analyse only an identified set of problems as

listed below.

9

List of Environmental Concerns

Air Water Solid waste Particulates (Total / Respirable) Microbial contamination Access to sanitation Indoor air pollutants Heavy metals Proximity to solid waste dumps SO2 NO2 Lead

Methodology

Population exposures to select air and water pollutants were assessed using a combination of

secondary data sources as well as primary sampling. Health risks were then assessed using

dose - response information established specifically from developing country studies. In

addition, primary data on the prevalence of respiratory, gastro-intestinal and vector-borne

diseases within the resident population of the study zone was also collected by administration

of a health assessment questionnaire. To assess economic costs of the health damage, local

information on costs of hospital visits, treatment and work-loss days specifically attributable to

environmental exposures were collected. Finally, the health and economic risks associated

with each environmental problem were assessed on the basis of data collected from the

preceding steps as well as on the basis of public perceptions and ranked accordingly.

Results

Air Quality

Air quality information was collected from secondary sources as well as through primary

sampling. Secondary sources included the National Ambient Air Quality Monitoring Database

(NAAQM) and Environmental Impact Assessment (EIA) reports of area industries done over a

period of last 5 years. Primary sampling was carried out for PM10 (Particulate matter less

than 10 µm), CO, SO2, NO2, lead and indoor air-pollutants (in houses using bio-fuels) through

a combination of personal sampling and area sampling techniques.

The environmental air quality data analyses revealed that levels of PM10 are the single

biggest concern. The results of limited primary sampling show that a significant fraction of the

10

population is exposed to PM 10 levels not adequately reflected in the area average reported

by the NAAQ database. Populations that use bio-mass fuels in homes for cooking or live in

slums adjacent to high traffic corridors, commuters, traffic police all represent categories that

are exposed to concentrations well in excess of the standards. Population exposure profiles

show that nearly 95% of the population is exposed to concentrations in excess of the World

Health Organisation (WHO) guideline values for PM10.

Annual 24-hour averages of lead in air as established through high volume sampling were

below the prescribed standards. However, blood lead levels were higher than the action level

of the USEPA. Since the relative contribution from air borne lead could not be ascertained

risks from elevated blood lead levels were separately calculated. Although the annual 24-hour

averages of CO, SO2 and NO2 were below the standards, the short-term exposure limits was

exceeded significantly during bio-mass burning while cooking or open refuse burning. The

limits were also exceeded during sporadic releases by the industry. Based on the above

results, health risk calculations were done only for PM 10 and for lead. Although the short-

term exposure limits were exceeded for other pollutants, the consequent health risks could

not be quantified due to the uncertainty in the dose -reponse relationship data for such

exposures.

Water Quality

Data on physico-chemical and microbiological parameters were collected predominantly from

the Ground water Cell of Chennai Metropolitan Water Supply and Sewerage Board

(CMWSSB). EIA reports of area industries were additional sources of information. Some

primary data was also collected, by sampling drinking water in select households. Secondary

data collected over a period of last five years revealed that microbiological contamination with

faecal coliforms was the most significant concern within the municipality. The data showed

presence of faecal coliforms in almost all sampling locations spread across the municipality

over multiple time frames.

Data on heavy metals including lead, chromium, arsenic, copper, iron and zinc showed that

levels were predominantly below the prescribed standards. Limited data on pesticides

showed no significant contamination of drinking water sources. Since the available data did

not show significant presence for any of the major chemical contaminants, health risk

11

calculations for water contaminants were limited to the assessment of prevalence of gastro-

intestinal disorders attributable to microbial contamination.

Solid waste & Access to Sanitation

Qualitative information on solid waste was collected both from municipal sources and from

household level surveys. Nearly 35% of the population did not have access to a private toilet

and made use of open grounds or public toilets. In slum populations nearly 80% did not have

access to private toilets. Around 55% of the population reported being severely affected by

rain- water stagnation. 30% lived in houses that were less than 100m meters from solid waste

dumps. The municipality was operating three of the solid waste dumps. Most of the dumps

were also found to contain large quantities of chemical and other hazardous wastes. The

access to these dumpsites was unrestricted. Rag pickers included many children from the

neighbouring slums. Improper solid waste disposal and lack of access to private sanitation

was cited as the most important environmental health concern in community perception

surveys. Quantitative health risk assessments were not performed for solid waste concerns,

as quantitative exposure information was not available. Prevalence of vector borne and water

borne infectious disease were collected from area hospitals as well as from study households.

Health risk assessment and economic valuation of health damage using previously established dose-response information

PM 10

Health End point Estimated Impact on study population*

Unit costs** (in Rs.)

Total costs (Lakhs of Rs.)

Pre mature deaths 202 393,879 797 Respiratory hospital admissions 285 1713 4 ERV (Emergency room visits) 5598 228.3 12 RAD/1000(Restricted activity days) 1367 41 563 RSD/1000(Respiratory symptom days) 4352 20 870 COPD (Chronic obstructive pulmonary disease) 1455 69,997 1018 LRI (Lower respiratory infections) in children 16,078 346 55 Asthma attacks 54275 200 108

TOTAL: 3432 LAKHS * Study population (covering an industrial municipality in North Chennai) - 211,000 **All costs are annual costs

12

Indoor air pollutants

Indoor air pollutants monitored included respirable particulates, carbon monoxide, SO2 and

NO2. Biomass combustion made the greatest contribution for exposures to these pollutants in

households (where in it was being used for cooking). The health risks from respirable

particulates were taken into account while reconstructing PM 10 exposures for the population.

Short term exposure guidelines for carbon monoxide was exceeded in 67% of homes during

cooking with bio-mass fuels while the guideline values for SO2 and NO2 were exceeded in

about 15% of the homes. Although gas exposures exceeded prescribed standards, the

consequent health risks could not be quantified due to the lack of definitive dose- response

relationships for short-term exposure.

Blood lead

Although lead levels in air and water were below the permissible standards, the blood lead

levels were still higher than the action level set by the USEPA. The mean blood lead levels in

children were 22µg/dl while in adults it was 16 µg/dl. Applying dose response information

cited in the main report, the calculated health risks for the study zone population were as

follows.

Mean Blood lead levels

Health End point

Estimated Impact on study population

Loss of IQ points 2 points/ child Children- 22µg/dl Increase in infant deaths 4.5 deaths /year

Increase in Blood pressure Men

Women

2.6 - 3.2 mm Hg 1.6 - 1.8 mm Hg

Heart Attacks 114/year Strokes 14/year

Adults - 16 µg /dl

Premature deaths 110/year

The economic costs associated with these health end points have not been computed. Unlike

PM10, studies on lead have come only from the developed country settings wherein the

doses are nearly an order of magnitude less than what has been observed in this study.

These risk calculations are therefore uncertain. Also, the cost of these health endpoints could

not be ascertained in this study and as there are no national figures available, a cost

estimation for these risks was not carried out.

13

Microbial contaminants in water and sanitation related issues were not examined using the

dose- response framework. Instead their contribution to health impacts was assessed from

the cross-sectional health and economic information gathered through the household survey

and the health information available at the local health care facilities.

Health impact assessment and economic valuation from cross-sectional epidemiological information gathered through the household survey

Prevalence of respiratory, gastrointestinal and vector borne illness was assessed in the study

zone through the administration of a household level questionnaire that collected both the

health and economic information. Treatment costs, wage loss days and defensive

expenditure attributable to specific health end points associated with air, water or sanitation

related issues were assessed to value the associated economic loss. Although, this method

was media specific (i.e. air/ water/solid waste) and not pollutant specific , this allowed

comparisons to be made between estimates that were obtained from applying previously

established dose-response functions and those obtained through direct assessment methods.

The economic costs of respiratory illness as determined from the house - hold survey is in

broad agreement with the economic costs for respiratory illness calculated using the dose-

response information for PM 10 exposure. Also the preceding calculations show that while

the health costs associated with air pollution are high, they are outweighed by water and

sanitation related concerns in the municipality.

Health end point

Prevalence Treatment costs

(lakhs of rupees)

Wage loss

(lakhs of rupees)

Defensive expenditure

(lakhs of rupees)

Total annual costs for study zone (lakhs of

rupees) Respiratory

illness 12% 710 942 1652

Gastro-intestinal disorders

16% 1167 1111 85 2363

Vector- borne illness

5% 1175 338 243 1757

14

Relative ranking of environmental concerns on the basis of health and economic risks

The health and economic risks obtained from the preceding steps were ranked relatively as

follows after taking into account the uncertainties surrounding these estimates as well as

people's perception (that was assessed as a part of the household survey).

Microbial contamination of water PM 10

High risks

Sanitation and solid waste disposal Lead Medium risks Indoor air pollutants from bio-mass combustion NO2 Low risks Volatile organics CO SO2 Chemical contaminants in drinking water

Uncertain risks

Hazardous wastes Conclusions

The study has been one of the first exercises within the city of Chennai, that has used the

quantitative health risk assessment framework to value environmental health damage.

Exposure assessment is one of the most crucial components for conduct environmental

health risk assessments and the present study through a combination of primary and

secondary data collection has generated a large database of population exposure information

for several key pollutants in the study zone. Although uncertainties still surround the

application of the dose-response information as well the use of cross sectional

epidemiological information, the study has clearly identified the need for initiating remedial

action for environmental concerns listed in the high risk category. The scheme of relative

ranking of risks allows one to prioritize between the concerns without the need to be

concerned about the absolute value of these risks. The economic valuation of health damage

assessed in this study can easily be used in subsequent cost-benefit analyses to choose

between alternate methods for reducing population exposures.

Most importantly the study has been executed with significant stakeholder participation

including members of the Tiruvottiyur Municipality, The Tamil Nadu State Pollution Control

Board and The Chennai Metropolitan Development Authority. The involvement of these

15

members will greatly facilitate the subsequent design and implementation of an environmental

management plan for the region. Institutional capacity building exercises in the form of

training in health risk assessment methodologies were carried out twice during the course of

this study. This would serve to further strengthen the local network of professionals to refine

these estimates for this zone as well as undertake similar exercises that would cover the

entire Chennai Metropolitan Area.

Finally the study has identified areas, where the largest gaps in data accuracy are likely to be

present. Recognition of these areas will contribute to refining the framework for conduct of

subsequent longitudinal environmental health studies. This will in turn ensure that the actual

health and economic costs are captured with the least degree of uncertainty for local policy

interventions. In that sense, it is hoped that the study will serve as good proto- type for other

studies in the region to recognize the importance of environmental health risks and initiate

action plans aimed at protecting the health of the public.

16

Background

Metropolitan cities across the world have a multitude of environmental concerns that affect

human health. Burgeoning populations, worsening air quality, polluted rivers, congested

roads, and hazardous occupational environments all represent problems faced by

environmental managers of all countries. Understanding the linkages between environment

and health is now well recognised as a key determinant of improved decision making during

the design of sustainable development plans for the regions. Addressing these issues in

developing country settings is all the more challenging as these compete with each other for

limited resources allocated for environmental improvement. A crucial task for environmental

decision-makers is therefore to decide on what would be the most efficient way of resource

allocation that would provide the most benefit to the concerned communities. Since the Rio

Earth Summit in 1992, the World Health Organisation (WHO) has augmented efforts for

understanding the environmental contribution to global ill health. The WHO in the 1997 report

on " Health and Environment and Sustainable Development ", made the first estimates of the

environmental contribution to the global burden of disease. These indicate that roughly 25-

33% of the global burden of disease is attributable to environmental factors with children

under 5 bearing the largest environmental burden. Figures 1 shows the environmental

contribution for all categories of diseases responsible for at least 1% of the total global burden

of disease (expressed as disability adjusted life years or DALYs*). The distribution of DALYS

in the major world regions is shown in Figure 2.

While the available statistics clearly identify the need for addressing environmental concerns

within development plans, regional environmental management plans have to rely on direct

assessments of the impact of local environmental degradation. In recent years, quantitative

health risk assessment and comparative risk assessment (CRA) procedures have been used

to quantify human health risks associated with particular environmental exposures and rank

them according to the magnitude and severity of health damage. These estimates can then

be subsequently used for assessing the economic impact of such damage. Together, they

can serve as the basis for the design of an environmental management plan that will integrate

this information with resource availability and assign priorities for allocation of resources.

The CRA process was originally developed by the United States Environmental Protection

Agency (USEPA) to determine how environmental priorities, based on health risks, lined up

17

with the level of resources devoted to mitigating problems within the EPA’s mandate (USEPA,

1989). The CRA approach is being used now, not only in the United States, but also in

several cities throughout the world, for the design of environmental management programs.

Two studies of this kind have been completed in India recently, in Ahmedabad and in Asansol

(USAID, 1995). A list of recent studies and environmental rankings arrived in these studies

are listed in Table 1.

While procedures for environmental health impact assessment have been used extensively to

quantify health benefits, the benefit may be viewed in terms other than health consequences.

The World Bank (Carter Brandon, 1995) has initiated such efforts, whereby the economic

costs of environmental degradation have been estimated for the several cities in India.

Economic costs associated with health impacts and productivity impacts have been

calculated, for several priority areas identified under the Government’s National

Environmental Action Programme (NEAP1993). These estimates give a good overview of the

issues and the costs at a national level and serve to demonstrate, qualitatively, the usefulness

of such an approach for setting of environmental priorities. Key economic estimates from

these studies for major Indian studies are summarised in Tables 2 and 3.

The health and economic estimates have been used to perform cost-benefit analyses for

interventions as well. The recent set of URBAIR documents published by the World Bank

present city specific data on air pollution as well as a framework for action plans for air

pollution control showing the viability of many of the technical control options.

For a rigorous cost benefit analysis of proposed environmental interventions in a region, it is

necessary that both the health impact assessments and the economic assessments be well

quantified. In the earlier CRA efforts in developing countries, the lack of accurate population

exposure information has resulted in considerable uncertainties in the health risk estimates,

which in turn cloud the economic damage estimates. These have been largely the result of

inadequate information in the available environmental and health databases. The existing

secondary data sources often do not have data amenable to apply the quantitative health risk

assessment framework. Further, in the absence of local information on the linkages between

the levels of exposures to particular pollutants and their health impacts (dose-response

information), often, developed country dose-response functions are applied to estimate the

magnitude of environmental health damage in populations of developing countries.

18

In India efforts by local institutions aimed at economic valuation of environmental health

damage have been few and far in between. A few studies that have been taken place have

focussed on Delhi and Mumbai. Metropolitan cities in Southern India have not applied a risk

assessment framework to begin addressing environmental concerns. It is with this

background that the present project was conceived in the city of Chennai. The present project

sought to demonstrate the usefulness of the CRA approach by addressing many of these

deficiencies through rigourous primary and secondary data collection on environmental,

health and economic parameters within one municipal zone of Chennai.

The Chennai metropolitan area is the fourth largest metropolis in India. Urban development

has been rapid over the last two decades. Much of the development has been through World

Bank and IDA supported initiatives under the Madras Urban Development Projects and the

Tamil Nadu Urban Development Project. The Chennai Metropolitan Development Authority

(CMDA) has been the coordinator for most of these efforts. Efforts to ensure sustainable

development have been initiated through programmes under the Sustainable Chennai project

of the United Nations Development Programme. The project has identified several strategies

for urban growth and development that would be compatible with environmental preservation.

Under the Sustainable Chennai project an exhaustive environmental profile has been

prepared that identifies several areas of environmental concern within the city (Appasamy,

Paul 1997). The concerns are based on review of the prevailing conditions in the city, in

terms of provision of essential services to the public, the civic infrastructure, environmental

pollution levels and other indicators that are likely to be detrimental to the quality of human

life. The issue of the impact of the prevailing environmental conditions on human health has

not been directly addressed.

The proposed project represents an effort, whereby the environmental health impacts have

been assessed through a combination of using previously established dose-response

information in conjunction with local exposure information as well as through collection of

primary environmental, health and economic information. This allowed cross-comparisons to

be made between the predicted and observed estimates as well as address the relative

uncertainties. The study represents one of the first studies in India to use a quantitative health

risk assessment framework to address environmental health concerns at a municipal level

with direct stakeholder involvement

19

Table 1: Recent Comparative Risk Assessment Studies

Study Ranking Highest Risk Air pollution from mobile sources High Risk Ambient Air Pollution Indoor Air Pollution Moderate Risk Drinking Water and Waste Water Low Risk Food Contamination Occupational Hazards Traffic Hazards No Rating Solid waste

Comparative Environmental Risk Assessment for Ahmedabad, India USAID (1995)

Particulate Matter Air Pollution Lead (in all media)

Higher Risks

Microbiological Diseases From Environmental Causes Microbial Food Contamination Middle Risks Ozone Air Pollution Sulphur dioxide Air Pollution Carbon Monoxide Air Pollution Indoor Air Pollution Drinking Water Contamination by Chemicals Drinking Water Contamination by Microbiological Agents

Middle/Lower Risks

Solid and Hazardous Wastes Toxic Air Pollutants Lower Risks Other Water Pathways: Direct Contact, Irrigation, Fish Consumption Nitrogen oxides air pollution Metals in Foods

Comparative Environmental Risk Assessment for Cairo, Egypt USAID (1994)

Uncertain Risks

Pesticides in Foods Indoor Radon Indoor air pollution other than radon Pesticides (primarily residues on food)

High Health Risks

Drinking water contamination Underground storage tanks Active hazardous waste (RCRA) sites Abandoned hazardous waste (CERCLA) sites Non hazardous (solid) waste sites

Comparing Risks and Setting Environmental Priorities Overview of Three Regional Projects USEPA (1989)

Low Health Risks

Nonpoint source discharges to surface waters

Table 2 : Estimates of Annual Health Incidences in Indian Cities due to Ambient Air

Pollution Levels exceeding WHO Guidelines

Cities Premature deaths Hospital admissions and sicknesses requiring medical treatment

Incidence of minor sickness (including RADs and RSDs)

Ahmedabad 2,979 1,183,033 72,177,644 Bangalore 254 135,887 8,326,282 Bhopal 663 277,847 17,024,691 Bombay 4,477 2,579,210 156,452,916

20

Calcutta 5,726 3,022,786 179,479,908 Delhi 7,491 3,990,012 241,958,219 Hyderabad 768 420,958 25,177,173 Jabalpur 683 286,214 17,537,333 Jaipur 1,145 520,947 31,708,958 Kanpur 1,894 812,381 49,247,224 Madras 863 461,966 27,859,487 Patna 725 319,242 19,561,109 Shimla 32 18,160 1,112,754 Shillong 0 0 0 Varanasi 1,851 785,414 48,125,143

Table 3 : Annual Costs of Ambient Air Pollution Levels in Indian Cities Exceeding WHO Guidelines, in US$

Cities Premature deaths Hospital admissions and

sicknesses requiring medical treatment

Incidence of minor sickness (including RADs and RSDs)

Lower estimate ($)

Upper estimate

($)

Lower estimate ($)

Upper estimate ($)

Lower estimate ($)

Upper estimate ($)

Ahmedabad 12,537,800 119,225,175 1,432,532 2,875,979 19,336,059 26,239,527 Bangalore 1,068,293 10,158,669 161,452 322,291 2,230,573 3,026,944 Bhopal 2,789,222 26,523,430 330,119 658,987 4,560,837 6,189,172 Bombay 18,839,236 179,147,155 3,201,644 6,474,425 41,913,018 56,877,036 Calcutta 24,094,972 229,125,302 4,088,226 8,462,469 48,081,843 65,248,289 Delhi 31,523,529 299,765,367 4,959,218 10,032,288 64,819,496 87,961,712 Hyderabad 3,230,756 30,722,096 553,527 1,137,350 6,744,849 9,152,932 Jabalpur 2,873,210 27,322,095 340,060 678,830 4,698,171 6,375,538 Jaipur 4,819,636 45,831,164 637,250 1,283,189 8,494,684 11,527,504 Kanpur 7,971,626 75,804,248 1,011,128 2,046,290 13,193,105 17,903,381 Madras 3,629,785 34,516,561 587,568 1,196,394 7,463,429 10,128,063 Patna 3,050,326 29,006,340 379,302 757,166 5,240,331 7,111,264 Shimla 133,985 1,274,093 21,577 43,072 298,102 404,532 Shillong 0 0 0 0 0 0 Varanasi 7,789,537 74,072,712 933,177 1,862,814 12,892,505 17,495,458 Pollutants : Particulate matters (PM10); Sulphur dioxide (SO2); Nitrogen oxides (NO2) Source: Carter Brandon and Kirsten Homman, The cost of inaction: valuing the economy – wide cost of environmental degradation in India. World Bank working paper,1995.

SCOPE OF PRESENT STUDY

The principal focus of the project was on assessing human health risks associated with

particular environmental exposures using both previously established dose – response

information and cross-sectional epidemiological information collected during the course of the

21

study. The economic costs associated with these health risks were then evaluated using local

economic information to enable ranking of the particular environmental concerns on the basis

of both the health and economic risks. The geographic focus of the project was the Thiruvottiyur Municipality of Chennai, located in the industrial corridor just outside the northern city limits, along the coast of the Bay of Bengal. Since even within a single zone,

there are a multitude of environmental concerns, the project sought to analyse only an

identified set of problems as listed below.

List of Environmental Concerns

Air Water Solid waste

Particulate Matter (Total / PM 10*) Microbial contamination Access to sanitation Indoor air pollutants Heavy metals Proximity to solid waste dumps SO2 NO2 Lead

* PM 10 - Particulate matter less than 10 µm in aerodynamic diameter.

Population exposures to select air and water pollutants were assessed using a combination of

secondary data sources as well as primary sampling. Health risks were then assessed using

dose - response information established specifically from developing country studies. In

addition, primary data on the prevalence of respiratory, gastro-intestinal and vector-borne

diseases within the resident population of the study zone was also collected through the

administration of a health assessment questionnaire. To assess economic costs of the health

damage local information on costs of hospital visits, treatment and work-loss days specifically

attributable to environmental exposures were collected. Finally, the health and economic risks

associated with each environmental problem was assessed on the basis of data collected

from the preceding steps as well as on the basis of public perceptions and ranked

accordingly.

The core team of investigators was drawn from the Sri Ramachandra Medical College &

Research Institute. In addition the study involved members from other academic and

governmental organisations including the Madras School of Economics, the Tamil Nadu

Pollution Control Board (TNPCB), the CMDA and most importantly the members of the

Thiruvottiyur Municipality. Since these Governmental bodies will be responsible for the

22

implementation of an environmental management plan at a later date, their participation was

sought to facilitate translation of the recommendations of the study into an action plan.

Finally, the project sought to specifically strengthen the technical capacity of institutions in

environmental health risk assessment and economic valuation procedures. Although not

explicitly listed in the original scope of the project, the investigators have developed a training

module on environmental epidemiology using the insight gained by the execution of this

project. The module has been administered to scientists and engineers of the regional

pollution control boards twice during the project period.

OBJECTIVES

Overall objective of the study

To rank a set of environmental problems in the Thiruvottiyur Municipality located within

the Chennai Metropolitan area on the basis of human health risks and the economic

impact of the health risks.

Specific Objectives of the study

To obtain available secondary environmental monitoring data for air and water

To conduct primary sampling and analyses for pollutants that are not routinely

monitored

To assess the prevailing status of solid and hazardous waste disposal

To assess population exposures for pollutants based on primary and secondary

environmental data

To assess health risks for select pollutants on the basis of previously established dose-

response information

To value the calculated health risks using local economic information

23

To collect information on prevalence rates for health end points such as respiratory

ailments, gastrointestinal and vector borne diseases that are specifically attributable to

environmental exposures, through the administration of a household level questionnaire

To collect information on health expenditure and wage loss, specifically associated

with health end points described above and thereby compute the specific economic costs.

To compare costs computed through the direct household level survey and those

estimated using predicted health risks

To collect information on community perceptions about relative magnitudes of

environmental problems of the zone

To rank the environmental problems on the basis of information collected in the

preceding steps

To provide the framework for the design of an environmental management plan for the

zone

To strengthen institutional capacity for CRA analyses through administration of training

METHODOLOGY

Overall framework of data collection

The population of the study zone is distributed across 48 municipal wards. The municipality,

for the provision of public health services has established five health posts. Since an essential

focus of the present project has been to demonstrate linkages between health and

environment and much of the secondary health data was available at the health post level,

the entire framework of data collection was centered around the health posts. Socioeconomic,

environmental and health data collected was organised under the areas covered by each of

the five health posts. Since the wards covered by each health post differed somewhat in

environmental quality this allowed cross comparisons to be made within the municipality. The

sequence of data collection proceeded along paths outlined as part A and B simultaneously

and were merged in Part C to make the final rankings.

24

Collection of Secondary Environmental Data Collection of primary environmental data Reconstruction of Population Exposures Health Risk assessment using previously established dose -response relationships Collection of local economic information on health costs Economic valuation

Collection of Secondary Health Data (From area health posts, hospitals, private medical practitioners etc.) Collection of Primary health data (Self reported using health assessment questionnaire) Assessment of prevalence of health end points attributable to environmental exposures Collection of health expenditure information from household level survey Economic valuation

Part A Part B Part C

Comparison of economic estimates obtained using the two approaches Assessment of relative uncertainties of the two approaches Assessment of community perceptions Relative ranking of environmental concerns

25

Part A

Collection of Secondary Environmental Data

For assessment of environmental exposures, all secondary data sources were covered

first prior to assessing the need for primary data collection.

Secondary data sources for air quality included the National Ambient Air Quality (NAAQ)

data available with the Tamil Nadu Pollution Control Board and the National

Environmental Engineering Institute, Chennai and environmental impact assessment

(EIA) reports conducted for the industries situated within the municipality over the

preceding 5 to 7 years. NAAQ data provided information on annual and monthly average

concentrations for suspended particulate matter, respirable particulate matter, sulphur-

dioxide and nitrogen dioxide. Occasionally information on lead concentrations was

available. EIA reports usually gave 24-hour averages of the above pollutants. In

addition data such as hydrocarbon levels, lead levels and carbon monoxide levels were

also available from few reports. Ozone, volatile organics and indoor air pollutant data

was not available in any of the secondary sources.

Secondary data sources for water quality included the data from the groundwater cell of

Chennai Metropolitan Water and Sewage Board (CMWSSB) and EIA reports of area

industries. The groundwater cell data contained information on physico-chemical and

bacteriological parameters while EIA reports provided additional information on heavy

metal concentrations as well.

Information on solid waste collection and disposal was collected from the municipality

office. Access to sanitation was assessed both on the basis of information provided by

the municipality in terms of wards connected through municipal sewer lines as well as on

the basis of information provided by the household level survey.

26

Collection of primary environmental data

Air

A preliminary analysis of secondary data was carried out which identified the areas

where available information was inadequate to estimate population exposures.

Secondary air data from the NAAQM and EIA reports revealed that concentrations of

SO2 and NOX were consistently below the prescribed National standards (National Air

Quality standards shown in Annexe 1). However, data on PM 10 concentrations showed

the area annual average reported by the NAAQM stations may not be truly reflective of

average population exposures. Several EIA reports showed PM 10 concentration to be

significantly in excess of the prescribed 24- hour standards. Although annual averages

were not computed in EIA reports, the figures pointed out to the possibility of

underestimating population PM 10 exposures using a single area average value that

was reported in the NAAQ database. Additional primary sampling was carried out for PM

10 to improve the secondary data estimates.

Primary sampling was carried out mostly through personal sampling (Reference

Procedure described in Annexe 1) using battery operated pumps that draw air through

impactors or cyclones designed to capture particles less than or equal to 10µm in

diameter. A few area samples using high volume samplers were also taken. Personal

samplers were attached to select groups that are likely to receive excessive exposures.

These included traffic policeman, slum dwellers, commuters who spent more than 8

hours on the roads and women cooks using bio-mass fuels for cooking. The population

exposures was reconstructed by taking the base exposure from the annual averages

reported in the NAAQ data and adjusting exposures of the high risk groups listed above,

upward on the basis of personal sampling data.

Similarly primary sampling was conducted through real time, direct read out data-logging

electrochemical sensors for carbon monoxide, sulfur - dioxide and nitrogen -dioxide.

These were done mostly in homes that used bio -mass fuels for cooking. They were also

used in traffic junctions to assess if the exposures exceeded the prescribed short -term

guidelines.

27

Monitoring for indoor air pollutants associated with combustion of bio-mass fuels in

homes using them for cooking was carried out as described above using personal

samplers for assessment of PM-10 exposures and direct read out instruments for gas

exposures. Personal samplers were attached to cooks or were placed near the cooking

spots both while cooking was going on and at other times. 24-hour personal exposures

for PM 10 were reconstructed on the basis of the area and personal measurements

using the time- activity information gathered from the members of the household. A

schematic for measurement locations and times is shown in Annexe 1. Gas

measurements were taken only while cooking was going on as at other times the

concentrations were below the threshold of detection. Indoor air quality monitoring was

conducted in a total of 86 households spread across the five

A limited set of measurements was carried out for volatile organics at traffic junctions

using a portable IR Analyser (MIRAN Sapphire supplied by Foxboro).

Water

Water samples were taken from drinking water sources (tap/ well/tanker) in all 48 wards

of municipality. The measurements were repeated twice during the study period. The

samples were analysed for physico-chemical parameters, bacteriological quality and

heavy metals as per procedures laid out in CPEEHO Manual (Cited in References

section). Results are shown as average of the two sets of measurements. Locations for

sampling are shown in the result section along with the results of analyses.

Collection of blood lead samples

Since lead was identified as one of the key environmental concerns and data on air

concentrations as well as water concentrations were limited, it was decided to ascertain

population exposures directly by analysing for blood lead levels. Detailed procedures

are listed in Annexe 1. Briefly, venous blood was collected in heparinised lead free tubes

and analysed using atomic emission spectroscopy (Inductively coupled). Analysis was

carried out at the Regional Sophisticated Instrumentation Centre of IIT Chennai. A

total of 143 samples were collected (56 children and 87 adults). Since the procedure

involved collection of venous blood, samples were collected only from people in the

municipality checking into area hospitals for other routine clinical chemistry

28

investigations. Informed consent was obtained prior to collection. However, since

complete addresses were not available it was not possible to assess if the population

was randomly distributed across the municipality or if they were drawn from particular

sections that may have had high end exposures (e.g traffic policemen, people living in

the vicinity of contaminated soils etc.). Since no other baseline information was

available, these levels are taken as the best possible estimates of population exposure.

Methodology for health risk assessment and economic valuation using previously established dose- response relationships

Environmental pollutants from the generators reach the humans after transportation and

transformation through environmental media. Since direct measurements of health

effects from environmental causes are rarely available, the process of risk assessment

quantifies health effects based on established dose –response relationships (i.e.

relationship between a dose of pollutant received and the expected magnitude of heath

damage). The dose or human exposure is quantified on the basis of either extrapolating

from environmental concentrations or through direct measurements in human body

fluids. The USEPA that first developed this procedure for risk assessment, has

compiled the health risk information on most of the major environmental contaminants in

the form of an electronic database termed IRIS (Integrated Risk Information System).

Data from IRIS was used in conjunction with recent epidemiological studies that have

shown significant correlation between exposures and health effects. Specific dose

response functions used for the various categories of pollutants are shown below.

PM10

Health impacts from PM 10 exposure are divided into mortality (excess deaths) and

morbidity (excess illness). These estimates are derived from a series of time series

studies that have looked at changes in mortality and morbidity over time as the PM

levels change or on the basis of cross sectional studies that compare effects across

different cities that differ in PM levels. While most of these dose response relationships

have been derived from studies in the developed countries, Ostro et. al. (Ostro, 1994)

have compiled dose -response information with specific reference to developing

29

countries. Empirical associations have been verified in studies carried out in Jakarta and

Metro Manila that have PM 10 exposures similar to what is observed in this study. The

dose- response function used for health risk calculations are based on those published

in the above reference and listed below.

Dose - response functions for PM 10 ( based on Ostro, 1992 and URBAIR 1997)

The health impacts are defined on a per unit change of PM 10 concentration

(expressed as µg/m3) . For this study the PM 10 benchmark is taken as 41 µg/m3 based

on the WHO air quality guidelines (1986). Although the latest guidelines (WHO, 2000)

show no threshold for health effects for PM 10 exposures, the risks in this study are

calculated based on the previous guideline. The risks may be therefore interpreted as

what may be avoided by bringing the PM 10 exposure levels to the guideline value of 41

ug/m3 and not as all risks associated with the current PM 10 exposures.

Following are the dose- response functions used in the study

1. % change in mortality =0.00112*(PM10-41)*Population exposed* Crude mortality rate

PM 10 = annual average concentration of PM 10 expressed in µg/m3

Crude mortality rate used in study = .0076

Exposed population = Fraction of population exposed to particular concentrations

(derived from population exposure profile that was

reconstructed on the basis of primary and secondary

environmental data)

2. Respiratory Hospital Diseases (RHD) per year per 100, 000 persons

= 1.2*(PM10-41)*Exposed Population

3. No: of emergency room visits (ERV) per year per 100, 000 persons

=23.54*(PM10-41)*Exposed Population

30

4. Restricted Activity days (RAD) per person per year

=0.0575*(PM10-41)*Exposed Population

5. Respiratory symptom days (RSD) per person per year

=0.183*(PM10-41)*Exposed Population

6. Change in yearly cases of chronic bronchitis per 100000 persons

= 6.12*(PM10-41)*Exposed Population

7. Chronic bronchitis in children below 18 years

=0.00169*0.4*(PM10-41)*Population

A population restriction factor of 0.4 is applied, as approximately 40% of the study

population is under 18 years old.

8. Change in daily asthma attacks per asthmatic person

=0.0326*0.07*(PM10-41)*Population

A population restriction factor of 0.07 is applied as approximately 7% of the population

are estimated to be asthmatic.

Valuation of Health Impacts for PM 10

The valuation of the health impacts determined above was done, using local economic

information that was obtained from the household level survey. The local economic

information collected included average annual income, average daily wages, cost of

doctor's visits, cost of medication, costs of hospital admissions etc. Using this

information the following unit costs were computed for each of the following health end

points. The unit costs for each health end point used computed in the study on the basis

of these local costs is listed below.

31

Unit costs of health impacts for PM 10

1. Change in Mortality (Premature deaths)

Premature deaths were valued using the value of a statistical life (VSL) which is

estimated as discounted value of expected future income at the average age. If

the average age of population is 24 years and the life expectancy at birth is 62

years, the VSL formula

t= 38

VSL = Σ w/(1+d)t where w = annual average income = Rs.22,044.

t=0 d = discount rate= 5%

Using the formula VSL for the study zone was estimated at Rs.393,880.

2. Respiratory hospital diseases

The valuation was based on an average of 9.1 days of hospital stay for each

admission resulting in a wage loss of Rs.88.3/day and a cost of Rs.100/day for

hospital charges and medication.

The unit cost of respiratory health diseases was estimated to be Rs. 1713.

3. Emergency room visits

The valuation was based on one wage loss day at Rs.88.3/day, cost of

medication at Rs.100/- per visit and cost of transportation at Rs.40 per visit.

The unit cost of respiratory health diseases was estimated to be Rs. 228.

4. Restricted Activity days

The valuation was based on Ostro's (1992) calculation of 20 % work-loss (valued

at average daily wage) and 80% lower productivity (valued at one third of average

daily wage).

The unit cost of restricted activity days estimated to be Rs. 41.

32

5. Respiratory symptom days

The valuation was based on average value reported in the Mumbai URBAIR

study as no information on willingness to pay to prevent a respiratory symptom

day is available.

The unit cost of restricted activity days is estimated to be Rs. 20.

6. Change in yearly cases of chronic bronchitis

The average age at which people become chronically ill with bronchitis is 35

years. Average life expectancy at birth is 62 years. It has been estimated that this

results in 50 days of wage loss/year. Also, it results in hospital visits at .5 times in

a year for an average of 13.1 days /visit. Finally medication costs are also

involved every year. Based on an average daily wage of Rs 88, a cost of

hospitalisation of Rs.1300 per admission and a cost of Rs.1000 per year for

medication and applying a discount of 5%, the total cost for an average of 27

years works out to be Rs. 105,043.

The unit cost of a case of chronic bronchitis is estimated to be Rs. 105,043.

7. Chronic bronchitis in children below 18 years

The duration of bronchitis averages 13.2 days, and is valued at respiratory

symptoms day (Rs.20). In addition two days of a parent's restricted activity day

valued at Rs41 per day was used.

The unit cost of respiratory health diseases was estimated to be Rs. 346.

8. Asthma attacks per asthmatic person

The valuation was based on information collected specifically from the household

level survey on expenditure incurred on medication / wage loss by asthmatics.

The unit cost of an asthma attack was estimated to be Rs. 200/ attack.

33

Other air pollutants

Health risk calculations were not performed for other air pollutants including gases as

the levels were below the prescribed standards for annual averages. Although short-

term exposure limits were violated, dose - response information specific for short-term

exposures is not currently available.

Blood lead

Changes in health risks to men, women and children with an increase in blood lead

levels were computed using various dose-response functions. The Centers of Disease

Control (CDC, 1991) has estimated that an increase in maternal blood lead levels by 1

µg/dL during pregnancy will result in an increased risk of infant mortality of 10-4.

Schwartz (1993) has estimated that for an increase of 1 µg/dL blood lead levels in

children, a decrease of 0.25 IQ points can be expected.

Elevated blood lead levels have been linked to elevated blood pressure in adults

(Schwartz, 1992). A meta-analysis of several studies estimated that a 1.4 mm Hg

increase in blood pressure can be expected for a 1 µg/dL increase in blood lead, in adult

males. In women the effect of an increase in blood lead levels is similar to that on men;

with diastolic blood pressure increasing by 0.8 mm of Hg with a 1 µg/dL increase in

blood lead level. The final dose response functions are as follows.

Increase in infant mortality: 10 -4 /ug/dL decrease in blood lead level

Change in blood pressure (Men) =1.44( PbB1 - Pb B2)_

where B1 = current blood lead level

B2 = blood level desired

Change in blood pressure in women = .6*1.44( PbB1 - Pb B2)_

The desired blood lead level was taken as 10µg/ dL the action level set by the USEPA.

The change in blood pressure was converted to increase in cardiovascular accidents

34

such as heart attacks and strokes using the age specific dose -response functions

described in McGee and Gordon(1976), Shurtleff(1974) and USEPA (1987).

Water Contaminants

Microbial contaminants in water and sanitation related issues were not examined using

the dose- response framework. Instead their contribution to health impacts was

assessed from the cross-sectional health and economic information gathered through

the household survey and the health information available at the local health care

facilities. Since chemical contaminants levels were less than prescribed standards

(albeit established on the basis of scanty environmental monitoring data), risk

assessment was not performed for chemicals in water.

Part B

Collection of secondary health data

Initial survey of hospitals and health posts revealed that record keeping in these facilities

was extremely poor and therefore could not be used for estimating prevalence of health

endpoints attributable to environmental exposures. The Infectious diseases hospital at

Tondiarpet had available records of inpatient data. The hospital was the main referral

hospital for acute diarrheal diseases and other viral diseases including chicken pox,

measles and mumps. However, outpatient data was not available. Other area hospitals

including Stanley Medical Hospital, the Govt. peripheral hospital and the clinics attached

to the health posts had no records of outpatient visits. Some basic health statistics were

available at the Municipality office. The available health records therefore were not used

in estimating prevalence of various health end points but only to qualitatively assess the

relative degree of prevalence.

Collection of primary health data and economic valuation

Because of the inadequacy of the secondary health data, primary health information was

collected through the administration of a health assessment questionnaire at the

household level. The health assessment questionnaire focused only on recording

35

respiratory symptoms, gastrointestinal disease symptoms and vector borne illness. For

vector borne illness only physician- diagnosed illness was recorded.

A total of 759 households were covered through questionnaires. A total of 3024

individual people were administered the questionnaire. The complete questionnaire is

furnished in Annexe 1. The questionnaire recorded general socio-economic information

and self reported symptoms based on a two- week recall. The questionnaire for

respiratory ailments was patterned after the questionnaires set by the British Medical

Research Council and the American Thoracic Society. The questionnaire also included

information on gastrointestinal illness episodes and vector/sanitation related disease

episodes. Information on disease specific health expenditures was also recorded

including wage loss days for particular classes of ailments.

Also, the study team recorded visits to the area hospitals/health post clinics/ private

practitioners either by deploying project staff to record the physician diagnosis or have

the concerned physician fill out a form that recorded their diagnosis. This was done to

find out the degree of reconciliation between hospital/ doctors office recorded illness

episodes and what was assessed through the primary household level survey. Although

prevalence of illness in the community is usually much higher than what is recorded by

area health care facilities, collection of this data allowed an assessment of population

patterns in seeking healthcare.

Total environmental health costs for the municipality was estimated on the basis of the

prevalence of particular health end points, the costs of treatment, costs of wage loss

days and defensive expenditure. Costs of treatment included costs of doctor visits, costs

of medication and costs of hospitalisation. Defensive expenditure included items such as

boiling water, purchasing water filters and purchasing mosquito control devices.

RESULTS

DESCRIPTION OF STUDY AREA

The town of Thiruvottiyur is located adjacent to the Chennai Metropolitan area,

sandwiched between Bay of Bengal on the eastern side and the Buckingham canal on

36

the western side. A single railroad divides the town into two parts. Most of the industrial

and residential developments have taken place on the eastern side of the railway line.

The jurisdiction of the municipality extends over 21.42 sq. km. The economic base is

one of mixed nature with both organized industrial activities and numerous unorganized

small-scale operations. The population of Thiruvottiyur is estimated to be 2.11 lakhs as

per the 1999 census data.

Socioeconomic Profile

The population is distributed across 48 municipal wards. The municipality, for the

provision of public health services has established five health posts. Since an essential

focus of the present project has been to demonstrate linkages between health and

environment and much of the secondary health data was available at the health post

level, the entire framework of data collection has centered around the health posts.

Basic socio-economic information describing the municipality is furnished in the following

tables.

Table 1: Historical Decadal Growth Rate of Population in Thiruvottiyur

Year Population Decadal Growth Rate 1901 15,919 - 1911 9409 -40.9 1921 8100 -13.9 1931 10732 32.5 1941 13909 29.6 1951 22393 61.0 1961 37764 68.6 1971 82853 119.4 1981 134014 61.7 1991 168642 25.8

Source: Census of India, Tamil Nadu, 1971 & 1991

Table 2: Annual Population Growth in Thiruvottiyur YEAR POPULATION ANNUAL GROWTH RATE NET ANNUAL ADDITION 1991 168139 - - 1992 171202 1.82 3063 1993 174193 1.75 2991 1994 181525 4.21 7332 1995 186703 2.85 5178 1996 191275 2.45 4572 1997 195364 2.14 4089

37

1998 205019 4.94 9655 1999 211100 2.97 6081

Source: Mid Year Population, Thiruvottiyur Municipality, Thiruvottiyur, 1999.

Table 3: Population Distribution By Health Posts In Thiruvottiyur (1999) S. No. Name of the Health Post Population Distribution (%)

1 Ernavoor 46095 21.8 2 Sathangadu 43497 20.6 3 Thiruvottiyur 53238 25.3 4 Thiruvottiyur Kuppam 27465 13.0 5 Thangal 40805 19.3

Total 211100 100 Source: Mid Year Population, Thiruvottiyur Municipalilty, Thiruvottiyur, 1999.

Table 4: Literacy Rate in Thiruvottiyur Sex 1971 1981 1991 Male 65.0 72.2 75.9 Female 43.2 54.8 61.4

Source: Census of India, Tamil Nadu, 1971, 1981 & 1991.

Table 5: Sex Ratio in Thiruvottiyur Year No. of Females per 1000 Males 1961 938 1971 865 1981 916 1991 927 Source: Census of India, Tamil Nadu.

Table 6: Birth and Death Rate in Thiruvottiyur

1997 1998 Birth Rate 8.9 7.9 Death Rate 3.2 2.9

Source: Municipal Health Office, Thiruvottiyur Municipality, Thiruvottiyur, 1999.

Table 7: Occupational Structure in Thiruvottiyur ( in %) 1971 1981 1991 Type of Occupation

Male Female Male Female Male Female Main Workers 33.9 2.9 32.8 4.1 31.4 5.3 Cultivators, Agrl. Labourers & Allied Activities

2.3 .10 .12 - 1.6 .1

Household Industry and Other Industrial Workers

17.3 .4 .38 .31 12.7 1.4

Construction 1.3 .18 32.3 3.7 2.2 .4 Trade, Commerce, Transport & Communication

8.1 .9 .2 .1 8.1 1.1

Others 4.9 1.42 - - 6.7 3.0 Non-Workers 32.2 94.1 34.2 91.8 37.3 88.7

Source: Census of India, Tamil Nadu, 1971, 1981 & 1991.

38

Environment Profile

Water Supply

Presently, this municipality gets water from an integrated scheme of MMWSSB, Bore-

wells and open wells. About 82% of the total population is covered by all the water

supply schemes. However, only 24% of the population is covered by protected water

supply. The per capita supply is only 15 lpcd. The integrated scheme is operated by

MMWSSB and the other schemes are operated and maintained by Thiruvottiyur

municipality.

Sanitation

The municipality has an underground sewage system for one-third of the area. However,

well water is used for flushing by most houses with individual toilet facilities. Sanitary

waste-water is discharged to the septic tanks for most parts. In terms of percentage of

population covered through various means of sullage disposal, while cesspools cover

45%, earthen drains cover 39%, and the combined underground drainage system

covers 15-30% of the households. All sullage and night soil is disposed in the nearby

low-lying areas without any treatment. The operation and maintenance of this service in

the town is therefore very poor.

Solid Waste Disposal

About 150 metric tonnes of solid wastes are generated per day in the municipality from

the individual households, industrial and commercial centers of which only about 80-90

tonnes are removed daily. These wastes are mainly collected and removed by

sweeping. While the frequency of sweeping at major roads, bus stand, railway station,

temple street etc is high, most other areas are swept on alternate days, some even once

in three days. The municipality employs 155 Male workers and 164 Female workers for

the conservancy work. Of these, 261 are permanent workers. A small fleet of vehicles

has been deployed by the municipality for transportation of the wastes to the dump-

yards, but a poor road infrastructure prevents access to collection for a majority of

houses.

39

About 12 acres of open vacant land at Manali near the railway track has been

sanctioned for solid waste dumping and composting. While municipal solid waste is

dumped here everyday no efforts have been taken to segregate waste categories and

initiate the composting facility.

Roads

The municipality being located very close to the Manali industrial area and the Chennai

port the volume of traffic is very high along the main arteries. The municipality has a

road density of 2.96 km/sq. km and surface road length of 47%. Ennore Express Way

and Thiruvottiyur High Road record the maximum volume with 17,409 pcu/day and

35934 pcu/day respectively. Kamaraj Salai also takes 10,221 pcu/day because of

access restrictions at the port for containers transporting hazardous materials.

Establishment of container freight stations and warehouses in the adjacent

municipalities of Ennore and Manali, has resulted in a high percentage of vehicles plying

on the road to take multiple trips. Sea erosion has lead to the shrinking of the right of

way on Ennore Expressway for a length of about 2 Kms. Pedestrians, parking, high

levels of commercial activity and accelerated degradation due to continuous exposure to

the marine environment have resulted in the residents facing severe and frequent

congestion on the roads.

Industries

The industrial make up includes a mixture of fertilizer, pharmaceutical, heavy chemicals,

automotive manufacturing, heavy engineering industries. Further, municipality is located

adjacent to the largest electric thermal power plant in Ennore industrial complex. Some

of the keys environmental issues associated with these industries are as follows;

Industrial plants produce gaseous emissions, released in different ways

treated or untreated to the atmosphere. Fugitive emission of particulates from

sources such as fertilizer plants are common. Incineration and flaring operations

also release SOx, NOx, Chlorinated organics, Hydrocarbons through the stacks.

Fly ash generation from the Ennore power plant is substantial.

40

Industrial discharges also are a major source of ground and surface water

pollution. A large number of industrial plants generate significant quantities of

hazardous wastes in the form of sludges that are highly toxic (e.g., mercury

bearing sludges from chloralkali operations, acid sludges from petrochemicals,

lime sludges etc.). These sludges together with chemical by-products, off

specification chemicals and discarded chemical containers accumulate in large

quantities on industrial sites. Their storage is not done in a manner so as to

prevent entry into surface or ground water.

Discharges of fly ash slurry into the sea are carried out routinely and the

resultant damage to the marine ecology remains to be quantified.

Storage of large piles of fluoride containing gypsum wastes has resulted in

elevated fluoride levels in ground water near plants where such wastes are

stored.

Intrusion of sea - water into subsurface water as a result of excessive

withdrawal of ground water is prevalent over the entire municipality.

More than 50 red category industries are located in and around the

municipality. Despite the requirement for performing environmental impact

assessments, no single database was available with detailed air and water quality

information that would allow an assessment of temporal and spatial variations in

pollution loads for the municipality.

The basic socioeconomic and environmental profile information described above is

shown in the following set of maps.

Analyses of environmental data

Air

PM 10

As described in the methodology section, air quality information was collected from

secondary sources as well as through primary sampling. Secondary sources included

41

the NAQM programme administered by the CPCB and TNPCB and EIA reports of area

industries done over a period of last 5 years. Since most secondary data sources show

data fro Suspended Particulate matter, they were converted to PM10 concentrations by

taking PM 10 to be 55% of SPM concentrations. Examples of data obtained from the

secondary sources are shown in graphs A1- A10. The results from the EIA reports show

that averages are higher than what is reported in the NAAQ database. Therefore

primary sampling was carried out for PM10 through a combination of personal sampling

and area sampling techniques. The locations of primary sampling was decided on the

basis of information collected on traffic density, proximity to roads, proximity to dump

yards where refuse is burnt etc. In addition the contribution of biomass combustion to

PM 10 exposures was also addressed in the study by sampling households where these

fuels were being used for cooking. The primary and secondary environmental

monitoring data was compiled for each of the five health posts and used in conjunction

with relevant exposure information (e.g. whether using bio-fuels, living adjacent to roads

etc.) to generate the population exposure profile.

The results of primary sampling reveal that a significant fraction of the population is

exposed to pollutant levels not adequately reflected in the area average reported by the

NAQM database. Populations that use bio-mass fuels in homes for cooking or live in

slums adjacent to high traffic corridors, commuters, traffic police all represent categories

that are exposed to concentrations well in excess of the standards. The environmental

data analyses revealed that levels of PM10 are the single biggest concern and that

biomass combustion is a significant source of PM 10 exposure to residents of homes

using these fuels.

Reconstruction of population exposure profiles on the basis both primary and secondary environmental shows that nearly 95% of the study population is exposed to concentrations in excess of the World Health Organisation (WHO) guideline values for PM10.

Key environmental air quality data and locations for air quality monitoring are shown in

the next set of figures.

42

The population exposure profiles were used to estimate health risks based on the

fraction of people at each exposure level and the dose –response relationships

described in the methodology section.

Lead

Annual averages of lead in air were available only for some years in the NAAQ database

and in a few EIA reports. All results indicated these levels to be below the prescribed

standards. However, blood lead levels were higher than the action level of the 10 µg/dL

prescribed by the USEPA. Since the relative contribution from air borne lead could not

be ascertained, risks from elevated blood lead levels were separately calculated. Risks

from air lead were not separately calculated.

Other air pollutants

Although the annual 24hour averages of CO, SO2 and NOX were below the standards,

the short-term exposure limits was exceeded significantly during bio-mass burning while

cooking or open refuse burning. The limits were also exceeded during sporadic releases

by the industry.

The levels of ozone and volatile organics were determined in areas with potential for

high exposures such as traffic junctions, through a combination of limited personal

sampling and area measurements. The 24-hour averages were below the standards.

However the measurements were too limited to make any deductions about annual

averages.

Based on the above results, health risk calculations were done only for PM 10 and for

lead. Although the short-term exposure limits were exceeded for other pollutants, the

consequent health risks could not be quantified due to the uncertainty in the dose -

reponse relationship data for such exposures.

Water

Data on physico-chemical and microbiological water parameters were collected

predominantly from the Ground water Cell of CMWSSB. EIA reports of area industries

were additional sources of information. Some primary data was also collected, by

43

sampling drinking water in select households. Secondary data collected over a period of

last five years revealed that microbiological contamination with faecal coliforms was the

most significant concern within the municipality. The data showed presence of faecal

coliformsin the range of 2- 2000 in all sampling locations spread across the municipality

and over multiple time frames.

Data on heavy metals including lead, chromium, arsenic, copper, iron and zinc showed

that levels were predominantly below the prescribed standards. Fluoride levels were

however significantly in excess in the adjacent Ennore municipality where large amounts

of fluoride containing gypsum waste were found stored. Limited data on pesticides

showed no significant contamination of drinking water sources.

Since the available data did not show significant presence for any of the major chemical

contaminants, health risk calculations for water contaminants were limited to the

assessment of prevalence of gastro-intestinal conditions in the population.

Solid waste & Access to Sanitation

Qualitative information on solid waste was collected both from municipal sources and

from household level surveys. Nearly 35% of the population did not have access to a

private toilet and made use of open grounds or public toilets. In slum populations nearly

80% did not have access to private toilets. Around 55% of the population reported being

severely affected by rain- water stagnation.30% lived in houses that were less than

100m meters from solid waste dumps.The municipality was operating three of the solid

waste dumps. Most of the dumps were also found to contain large quantities of chemical

and other hazardous wastes. The access to these dumpsites was unrestricted. Rag

pickers included many children from the neighbouring slums. Improper solid waste

disposal and lack of access to private sanitation was cited as the most important

environmental health concern in community perception surveys.

Quantitative health risk assessments were not performed for solid waste concerns, as

quantitative exposure information was not available. Prevalence of vector borne and

water borne infectious disease were collected from area hospitals as well as from study

households.

44

HEALTH RISK ASSESSMENT AND ECONOMIC VALUATION OF HEALTH DAMAGE

Health risk assessments using previously established dose response functions

Health risk assessments were done using both previously established dose response

information and prevalence information for respiratory, gastro-intestinal and vector borne

problems gathered through the administration of a health assessment questionnaire.

Economic information collected included wage information; costs of medical treatment

(medicines/ consultation/ hospitalisation costs) for private and Municipal health care

facilities; average number of work loss days for specific ailments and associated wage

loss.

Economic valuation was done using this information and applying it to either the

calculated health risks or the prevalence information gathered through the primary

health survey.

Other studies have used the dose –response relationships to estimate economic costs.

As seen form the following tables, the costs tend to be higher when used with detailed

exposure information as opposed to using zonal averages provided by the NAAQ

database. This indicates that the available information may under represent both

population pollutant loads and consequently the associated health and economic costs.

The cost estimates for particular pollution loads are comparable between the present

study and the URBAIR study in Mumbai.

PM-10

Use of previously published dose - response information with the local economic information gave the following results (Based on population exposure profiles)

Health End point Impact on study Population*

Unit costs (Rs) Total costs (in lakhs of rupees)

Pre mature deaths 202 393,879 797 Respiratory hospital admissions 285 1713 4 ERV (emergency room visits) 5598 228 12 RAD/1000(Restricted activity days) 1367 41 563 RSD/1000(respiratory symptom days)

4352 20 870

COPD 1455 69,997 1529

45

Chronic bronchitis in Children 16,078 346 55 Asthma 54275 200 108

* Study population- 211,000 Total : Rs. 3432 Lakhs Comparison of economic estimates for health damage resulting from PM 10 exposures with other studies using same dose response information

Study zone (Using only NAAQ Data) Rs.865/person Study zone (Using both NAAQ and primary sampling data; includes population exposure profiles)

Rs.1626/person

Chennai Average (World Bank estimates using NAAQ data)

Rs275/person

Bombay Average (World Bank estimates using NAAQ data)

Rs.734/person

Bombay (URBAIR, World Bank estimates using modelling+ monitoring data; includes population exposure profiles)

Rs.1840/person

Lead

Similar to PM 10, previously published quantitative dose-response information is

available for calculation of health risks from lead. Since lead enters the human system

through several routes, including air water and soil, blood lead is usually considered the

best surrogate for total exposure. Further, in the study zone, average lead levels in air

and water were below the national standards and therefore could not be used for

calculation of health risks. Blood lead measuremen